Orthosiphon stamineus (Misai Kucing) for Diabetes: A Translational Reappraisal of Human Evidence, Dose, Mechanistic Assays and Safety

1. Introduction

Orthosiphon stamineus Benth. (synonym O. aristatus), known as misai kucing in Malaysia, kumis kucing in Indonesia and Java tea in Europe, is a Lamiaceae herb used in Southeast Asia and southern China for diabetes, kidney disorders and urinary complaints [1]. In Malaysia it is widely consumed as a herbal tea and used in many herbal products.

A 2022 systematic review by Wang and colleagues included 31 studies of O. stamineus in diabetes and its complications [1]. It described mechanisms including inhibition of α-glucosidase and α-amylase, antioxidant and anti-inflammatory activity, improved insulin secretion and sensitivity, increased glucose uptake and glucagon-like peptide-1 secretion, identified phenolic acids, flavonoids and triterpenoids as probable active constituents, and concluded that O. stamineus "could be an antidiabetic agent". The authors noted that clinical research was very limited.

This review takes up that limitation. It asks four questions the 2022 review did not address directly: what human evidence exists; whether the doses that worked in animals correspond to amounts people can take; whether the in vitro assays used to establish mechanisms are predictive; and what safety and interaction issues arise for people with diabetes, who usually take several medicines and often have kidney disease.

2. Methods

Europe PMC and Crossref were searched to October 2026 for human studies of O. stamineus or O. aristatus with any glycaemic, renal or diuretic outcome, and for pharmacokinetic, interaction and toxicity studies. The reference list of the 2022 review [1] was screened, and the draft European Union herbal monograph [2] was consulted. Effective doses from animal studies cited in the 2022 review were converted to human-equivalent doses using the body-surface-area method, dividing the animal dose in mg/kg by 12.3 for mice and 6.2 for rats [3], and expressed for a 60 kg adult. These conversions are approximate and are used only to compare orders of magnitude. This is a narrative critical review.

3. Human Evidence

No randomised controlled trial of O. stamineus on blood glucose, HbA1c or insulin sensitivity was identified (Table 1). The 2022 review cited two Chinese clinical studies in diabetic nephropathy, each of 30 patients, reporting "effective rates" of 76.7% and 83.3% [1,4,5]. Both used multi-herb prescriptions in which O. stamineus was combined with Cordyceps sinensis, Astragalus membranaceus, rhubarb, leech and other materials, so any effect cannot be attributed to O. stamineus.

The only placebo-controlled, double-blind human trial of O. stamineus identified tested its best-known traditional use. In a crossover study of four Vietnamese diuretic remedies, O. stamineus had no effect on 12-hour or 24-hour urine output or sodium excretion [6]. The European Medicines Agency's draft monograph accepts O. stamineus only as a traditional herbal medicine, based on long-standing use, to increase urine output for flushing the urinary tract, and does not list any metabolic indication [2]. The human evidence for glycaemic effects is therefore absent, and the evidence for the traditional diuretic effect is weak.

4. Dose Translation

Table 2 converts the effective doses in animal studies cited by the 2022 review into human-equivalent amounts. The mouse studies used doses equivalent to 0.8 to 3.3 g of extract a day for a 60 kg adult, within or close to the range of the European monograph, where the highest daily dose of a dry extract is 1.44 g [2]. These studies were in high-fat-diet or normal mice and measured fasting glucose, glucose tolerance or body weight [7,8]. The rat studies, which provided most of the evidence in diabetic models, used 0.5 to 2 g/kg/day, equivalent to 4.8 to 19.4 g a day, or 3 to 13 times the monograph maximum [9,10,11]. The highest of these used a chloroform sub-fraction, a small fraction of the original extract, so the equivalent amount of leaf would be greater still.

The relevant doses also depend on composition. Across 15 genotypes of O. aristatus, leaf rosmarinic acid content varied from 0.06 to 7.25 mg/g and sinensetin from 0.36 to 4.02 mg/g [12], 120-fold and 11-fold ranges respectively. Animal studies used aqueous, 50% ethanol, ethanol and chloroform extracts of materials of different origin, and the 2022 review did not tabulate their marker contents [1]. A product cannot be said to deliver an effective dose unless both the amount and the composition of the extract tested are known.

5. Mechanistic Assays

5.1. α-Glucosidase Inhibition

Inhibition of intestinal α-glucosidase is the mechanism most often cited and the one most plausible at oral doses, because it acts in the gut lumen and does not require absorption. The 2022 review reported that an O. stamineus fraction inhibited α-glucosidase more potently than acarbose (IC50 40 versus 250 µg/mL) [1]. In the study of a standardised 50% ethanol extract, the enzyme was described only by supplier, and acarbose had an IC50 of 1.93 mg/mL [13]. Acarbose is a potent inhibitor of mammalian intestinal α-glucosidases but inhibits the yeast enzyme weakly or not at all, whereas many food polyphenols inhibit the yeast enzyme while having little effect on the rat intestinal enzyme [14]. An acarbose IC50 in the milligram-per-millilitre range is consistent with a yeast enzyme. Comparisons in which plant extracts outperform acarbose in such assays do not predict effects in the human gut. The in vivo starch- and sucrose-loading studies in rats are more relevant [11], but they used doses equivalent to almost 10 g of extract a day (Table 2).

5.2. Absorption of the Proposed Active Constituents

Most of the other mechanisms in the 2022 review, including effects on insulin signalling, glucose uptake and inflammation, were shown in cell cultures at micromolar concentrations of extracts or pure compounds [1]. Whether such concentrations are reached in people is unknown for O. stamineus. Rosmarinic acid, one of its main markers, reached a peak serum concentration of 162 nmol/L in humans after a single 500 mg dose given as a lemon balm extract [15], and is largely metabolised by gut microbiota and conjugated [16]. A 500 mg dose of rosmarinic acid corresponds to roughly 70 g of the richest O. aristatus leaf in the genotype survey [12]. Sinensetin and other methoxylated flavones are absorbed but extensively demethylated, sulfated and glucuronidated in rats [8,17]. Mechanisms demonstrated with the parent compounds in cells should therefore be regarded as hypotheses until plasma or tissue concentrations after realistic doses are measured.

6. Safety and Interactions

The animal toxicology of O. stamineus is reassuring (Table 3). A standardised 50% ethanol extract caused no toxicity at up to 5000 mg/kg/day for 28 days in rats [18], an aqueous extract caused no developmental toxicity at up to 2000 mg/kg/day [19], and the aqueous extract was not genotoxic [20]. These findings support the safety of traditional tea use.

People with diabetes, however, usually take several medicines. In vitro, dichloromethane and petroleum ether extracts of O. stamineus inhibited CYP3A4, and the flavone eupatorin inhibited CYP2D6 with a Ki of 10.2 µM, whereas no significant inhibition of CYP2C9 was found [21]; effects on CYP2C19 have also been reported [22]. The inhibitory constituents are lipophilic flavones, so concentrated or ethanolic extracts may carry more interaction potential than aqueous teas. If O. stamineus does lower glucose, additive effects with insulin or sulfonylureas would be expected; no interaction study has addressed this.

A further concern arises from the population for which the herb is promoted. The 2022 review proposed clinical study of O. stamineus in diabetic nephropathy [1], and in China it is used for chronic renal failure. The European monograph contraindicates its use in conditions where reduced fluid intake is advised, such as severe cardiac or renal disease, and requires adequate fluid intake during use [2]. Patients with advanced diabetic kidney disease may be on fluid restriction, potassium-sparing or loop diuretics and sodium-glucose cotransporter-2 inhibitors. Any trial in this population would need careful monitoring of fluid balance, electrolytes and kidney function.

7. Implications

7.1. For Products and Claims

On current evidence, O. stamineus products should not carry claims to lower blood glucose or treat diabetes. The traditional use accepted by the European regulator is a urinary flushing indication, and even its diuretic effect was not confirmed in the one controlled human trial [2,6]. Products should state their extract type, drug-to-extract ratio and content of rosmarinic acid, sinensetin and eupatorin, so that dose and interaction potential can be judged. People taking glucose-lowering drugs or with kidney disease should consult their doctor before use.

7.2. For Research

The next step should be a randomised, double-blind, placebo-controlled trial of a standardised extract at a dose within the European monograph range, for example 0.8 to 1.2 g of dry extract a day, in adults with prediabetes or early type 2 diabetes on stable metformin or no medication. HbA1c after 12 to 24 weeks would be the primary outcome, with fasting and post-meal glucose, insulin sensitivity, and safety including kidney function and electrolytes. A meal-tolerance substudy could test α-glucosidase inhibition directly, and plasma concentrations of rosmarinic acid, sinensetin, eupatorin and their metabolites should be measured. In vitro work should use mammalian intestinal α-glucosidases and report marker contents, and a clinical interaction study with a CYP3A4 probe would establish whether the in vitro inhibition matters.

7.3. Limitations

This review is narrative. Human-equivalent dose conversions are approximate and do not account for differences in extract composition. Two Chinese clinical studies were assessed from the description in the 2022 review. The enzyme origin in some α-glucosidase studies was not stated and was inferred from the behaviour of acarbose. The author's company manufactures botanical extracts.

8. Conclusions

The 2022 systematic review established a broad preclinical rationale for O. stamineus in diabetes. Translated to people, that rationale is weaker than it appears. There is no randomised trial on glycaemic outcomes, the only controlled human trial did not confirm the traditional diuretic effect, the rat doses that lowered glucose correspond to several times the maximum regulatory daily dose, the α-glucosidase comparisons with acarbose were probably made with an enzyme that does not predict human effects, and the proposed active constituents are poorly absorbed and extensively metabolised. Animal toxicity is low, but lipophilic extracts may interact with drugs metabolised by CYP3A4 and CYP2D6, and use is contraindicated where fluid intake must be restricted. O. stamineus should be regarded as a traditional herbal tea with an untested glycaemic hypothesis until a standardised extract at regulatory doses is evaluated in a randomised trial.